Hans Kersting

gradient descent (3) stochastic differential equation (2) noise injection (2) saddle point (2) stochastic gradient descent (2) sharpness-aware minimization (1) minimax optimization (1) gaussian process (1) stochastic process (1) noise correlation (1) ornstein-uhlenbeck process (1) overparameterized model (1) flat minima (1) fractional brownian motion (1) gradient descent ascent (1) ito calculus (1) overparameterized neural network (1) explicit regularization (1) neural network generalization (1) layer-wise perturbation (1)